Data centers, such as Google's facility in Dalles, Oregon, generate huge amounts of waste heat.

Data centers, such as Google's facility in Dalles, Oregon, generate huge amounts of waste heat. Google

Waste Heat: Innovators Turn to an Overlooked Renewable Resource

Nearly three-quarters of all the energy produced by humanity is squandered as waste heat. Now, large businesses, high-tech operations such as data centers, and governments are exploring innovative technologies to capture and reuse this vast renewable energy source.

When you think of Facebook and “hot air,” a stream of pointless online chatter might be what comes to mind. But the company will soon be putting its literal hot air — the waste heat pumped out by one of its data centers — to good environmental use. That center, in Odense, Denmark, plans to channel its waste heat to warm nearly 7,000 homes when it opens in 2020.

Waste heat is everywhere. Every time an engine runs, a machine clunks away, or any work is done by anything, heat is generated. That’s a law of thermodynamics. More often than not, that heat gets thrown away, dribbling out into the atmosphere. The scale of this invisible garbage is huge: About 70 percent of all the energy produced by humanity gets chucked as waste heat.

“It’s the biggest source of energy on the planet,” says Joseph King, one of the program directors for the U.S. government’s Advanced Research Projects Agency-Energy (ARPA-E), an agency started in 2009 with the mission of funding high-risk technology projects with high potential benefit. One of the agency’s main missions is to hike up energy efficiency, which means both avoiding making so much waste heat in the first place, and making the most of the heat that’s there. ARPA-E has funded a host of innovative projects in that realm, including a $3.5 million grant for RedWave Energy, which aims to capture the low-temperature wasted heat from places like power plants using arrays of innovative miniature antennae. 

“The attitude has been that the environment can take this waste,” says one expert. “Now we have to be more efficient.”

The problem is not so much that waste heat directly warms the atmosphere — the heat we throw into the air accounts for just 1 percent of climate change. Instead, the problem is one of wastage. If the energy is there, we should use it. For a long time, says Simon Fraser University engineer Majid Bahrami, many simply haven’t bothered. “The attitude has been that the environment can take this waste; we have other things to worry about,” he says. “Now we have to be more efficient. This is the time to have this conversation.”

The global demand for energy is booming — it’s set to bump up nearly 30 percent by 2040. And every bit of waste heat recycled into energy saves some fuel — often fossil fuels — from doing the same job. Crunching the exact numbers on the projected savings is hard to do, but the potential is huge. One study showed that the heat-needy United Kingdom, for example, could prevent 10 million tons of carbon dioxide emissions annually (about 2 percent of the country’s total) just by diverting waste heat from some of the UK’s biggest power stations to warm homes and offices. And that’s not even considering any higher-tech solutions for capturing and using waste heat, many of which are now in the offing.

To help reduce carbon emissions — not to mention saving money and lessening reliance on foreign fuel imports — governments are increasingly pushing for policies and incentives to encourage more waste heat usage, big businesses like IBM are exploring innovative technologies, and start-ups are emerging to sell technologies that turn lukewarm heat into usable electricity. 


For more than a century, waste heat has been used for its most obvious application: heat (think of your car, which uses waste heat from its engine to heat your interior). In 1882, when Thomas Edison built the world’s first commercial power plant in Manhattan, he sold its steam to heat nearby buildings. This co-generation of electricity and usable heat is remarkably efficient. Today, in the United States, most fossil fuel-burning power plants are about 33 percent efficient, while combined heat and power (CHP) plants are typically 60 to 80 percent efficient

When it opens in 2020, Facebook's new data center in Odense, Denmark will channel its waste heat to warm nearly 7,000 homes.

When it opens in 2020, Facebook's new data center in Odense, Denmark will channel its waste heat to warm nearly 7,000 homes. Facebook

That seems to make co-generation a no-brainer. But heat is harder to transport than electricity — the losses over piped distances are huge — and there isn’t always a ready market for heat sitting next to a power plant or industrial facility. Today, only about 10 percent of electricity generation in the U.S. produces both power and usable heat; the Department of Energy has a program specifically to boost CHP, and considers 20 percent a reasonable target by 2030.

Other countries have an easier time thanks to existing district heating infastructure, which uses locally produced heat to, typically, pipe hot water into homes. Denmark is a leader here. In response to the 1970s oil crisis, the country began switching to other energy sources, including burning biomass, which lend themselves to district heating. As a result, Denmark has an array of innovative waste-heat capture projects that can be added onto existing systems, including the upcoming Facebook data center. 

In 2010, for example, Aalborg’s crematorium started using its waste heat to warm Danish homes (after the Danish Council of Ethics judged it a moral thing to do). Others are joining in. In Cologne, Germany, the heat of sewage warms a handful of schools. In London, the heat from the underground rail system is being channelled to heat homes in Islington. An IBM data center in Switzerland is being used to heat a nearby swimming pool. “Data centers crop up again and again as having huge potential,” says Tanja Groth, an energy manager and economist with the UK’s Carbon Trust, a non-profit that aims to reduce carbon emissions.

An alternative option is to turn waste heat into easier-to-transport electricity. While many power plants do that already, regulators striving for energy security are keen to push this idea for independent power producers like large manufacturers, says Groth. Businesses that make their own power would reduce carbon emissions by getting any extra electrical juice they need by squeezing it out of their waste heat, rather than buying it from the grid.

Waste heat is a problem of a thousand cuts, requiring a mass of different innovations.

Several companies have popped up to help do just this. One of the largest, Turboden, based in Brescia, Italy, sells a mechanical system based on the Organic Rankine Cycle. This is a type of external combustion engine — an idea that pre-dates the internal combustion engine used in cars. Rankine engines and similar technologies have contained, closed-loop systems of liquid that expand to gas to do work, thanks to a temperature difference on the outside of the system — so you can drive a power-generating engine off waste heat. When a cement plant in Bavaria, for example, added a Rankine engine to its system a decade ago, it reduced its electricity demand by 12 percent and its CO2 emissions by about 7,000 tons.

Since 2010, Turboden says it has sold systems for waste heat recovery to 28 production plants, with seven more under construction now. Turboden is just one of many; the Swedish-based company Climeon, for example, endorsed by spaceflight entrepreneur Richard Branson, uses a similar but different technique to make an efficient heat engine that can be bolted onto anything industrial, from cement plants to steel mills, in order to recycle their waste heat.


Waste heat is a problem of a thousand cuts, requiring a mass of innovations to tackle different slices of the problem: a system that works for one temperature range, for example, might not work for another, and some waste heat streams are contaminated with corrosive pollutants. “We aren’t looking for a silver bullet,” says Bahrami. “There are so many different things that can be done and should be done.”

Heat radiates from the Grangemouth Oil Refinery in Scotland. About 70 percent of all the energy produced globally gets discarded as waste heat.

Heat radiates from the Grangemouth Oil Refinery in Scotland. About 70 percent of all the energy produced globally gets discarded as waste heat. Christopher Furlong/Getty Images

Bahrami and others are pursuing solid-state systems for waste heat recovery, which, with no moving parts, can in theory be smaller and more robust than mechanical engines. There are a wide array of ways to do this, based on different bits of physics: thermoacoustics, thermionics, thermophotovoltaics, and more, each with pros and cons in terms of their efficiency, cost, and suitability to different conditions.

“Thermoelectrics have been the major player in this space for years,” says Lane Martin, a material scientist at the Univeristy of California, Berkeley and Lawrence Berkeley National Laboratory. Seiko released a “thermic watch” in 1998 that ran off the heat of your wrist, for example, and you can buy a little thermoelectric unit that will charge your cell phone off your campfire. Researchers are trying hard to increase the efficiency of such devices so they make economic sense for wide-scale use. That means screening thousands of promising new materials to find ones that work better than today’s semiconductors, or tweaking the microstructure of how they’re built.

The biggest technological challenge is to pull energy from the lukewarm end of the spectrum of waste heat: More than 60 percent of global waste heat is under the boiling point of water, and the cooler it is, the harder it is to pull usable energy from it. Martin’s group is tackling this by investigating pyroelectrics (which, unlike thermoelectrics, works by exploiting electron polarization). This isn’t near commercial application yet; it’s still early days in the lab. But the thin-film materials that Martin’s team is investigating can be tuned to work best at specific temperatures, while thermoelectrics always work better the larger the temperature difference. Martin imagines future systems that stack thermoelectric materials to suck up some of the warmer waste heat, say above 212 degrees Fahrenheit, and then uses pyroelectrics to mop up the rest. Martin says his recent work on such materials drummed up interest from a few bitcoin mining operations. “They have a real problem with waste heat,” says Martin. “Unfortunately, I had to tell them it’s a little early; I don’t have a widget I can sell them. But it’s coming.” 

“Waste heat is an afterthought — we’re trying to make it a forethought,” says a U.S. government scientist.

Perhaps one of the best applications for waste heat is, ironically, cooling. Air conditioners and fans already account for about 10 percent of global energy consumption, and demand is set to triple by 2050. In urban areas, air conditioners can actually heat the local air by nearly 2 degrees F, in turn driving up the demand for more cooling.

One solution is to use waste heat rather than electricity to cool things down: absorption or absorption coolers use the energy from heat (instead of electrically driven compression) to condense a refrigerant. Again, this technology exists — absorption refrigerators are often found in recreational vehicles, and tri-generation power plants use such technology to make usable electricity, heat, and cooling all at once. “Dubai and Abu Dhabi are investing heavily in this because, well, they’re not stupid,” says Groth.

But such systems are typically bulky and expensive to install, so again research labs are on a mission to improve them. Project THRIVE, led in part by IBM Research in Rüschlikon, Switzerland, is one player aiming to improve sorption materials for both heating and cooling. They have already shown how to shrink some systems down to a reasonable size. Bahrami’s lab, too, is working on better ways to use waste heat to cool everything from long-haul trucks to electronics.

It’s very hard to know which strategies or companies will pan out. But whatever systems win out, if these researchers have their way, every last drop of usable energy will be sucked from our fuel and mechanical systems. “Waste heat is often an afterthought,” says King. “We’re trying to make it a forethought.”